25 ms·

It has never been realistic for people to find the truth, perhaps save those who are focused on a very narrow subject of truth and are dedicating their lives to

It has never been realistic for people to find the truth, perhaps save those who are focused on a very narrow subject of truth and are dedicating their lives to the study. We have always relied on a collection of people with the aforementioned focus with some system of accreditation to tell us what is true. While far from perfect, I'm not sure a deluge of generated AI nonsense outside of the vetted channels, complementing people already stirring up nonsense, changes anything.

toss1 3y ago
>>It has never been realistic for people to find the truth, Rather strained definition of truth there. The truth that matters is a basic not-very-wrong version of what is happening. Examples: 1) The earth is roughly spheroidal. 2) NASA is a real space agency and landed people on the moon in 1969-1973. 3) The trails behind airplanes are condensation particles, not sprayed chemicals. 4) Vaccines are safe and effective, and the ratio of issues to prevented disease is on the order of millions-to-one, a vast benefit to society, especially those who medically cannot get vaccines. 5) Hijacked airlines and not pre-set explosives brought town the WTC towers in NYC on 9/11/2001. Etc. Etc. Etc.... We do NOT need a doctorate in Geology, selection into NASA Astronaut Corps, graduate astronomic physics course, PhD-level immunology, or a structural engineer's license to understand any of the above to a useful level. Yet, there is sufficient human-generated noise to convince millions that none of these are true. And you are claiming that adding industrial-scale AI agents to the public 'conversation' at a scale never seen in history will have zero effect? Sure, it might completely fizzle. But the normal and prudent expectation when something comes in and "floods the zone with shit" [0], especially when that shit has the potential to be at a volume that is far beyond anything experienced, is to expect that it will have some effect. What evidence do you have that it will have no effect? In answering, consider another top HN story at the moment "Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real". [1] [0] Steve Bananon's description of a disinformation tactic [1] https://news.ycombinator.com/item?id=38684309 https://news.ycombinator.com/item?id=38684309
- randomdata 3y ago
  > to understand any of the above to a useful level. The standard model for electrical current is not true, but it is still useful in the vast, vast, vast majority of cases. True and useful are very different things. The conflation here is erroneous. The average person is never going to understand what is true in any meaningful way, and even those who focus on certain truths are going to be focused on very specific truths, not all truths. There is only so much time in the day. For everything else they have to work with simplified models that may be quite disconnected from reality, but as long as they work for all practical purposes, who cares? > And you are claiming that adding industrial-scale AI agents to the public 'conversation' at a scale never seen in history will have zero effect? Conversation only happens when people realize they are working with a faulty model, so if conversation keeps happening then that means people are realizing that the model they currently have doesn't work. What do you think is going to happen with more noise? If people can't find a usable model in a not-so-noisy environment, do you think they will suddenly be able to find one in a noisier environment? That seems quite unlikely. In reality, the outcome is the same either way. Maybe increased noise will turn more interest to the vetted environments and ignoring the noisy conversations, but if people are already not finding the vetted, the reality is that we are not likely to see change there either.
  - toss1 3y ago
    >>The average person is never going to understand what is true in any meaningful way, Agree, but this means we are talking about two different topics. I'm talking at a functional level of truth; a base level of truths commonly understood in a society such that it can maintain cohesion. You're talking about deeply understood truths, like the 2+-hour discussion of quantum effects on the question of "why is the sky blue?" at a PhD's defense presentation. >> Conversation only happens when people realize they are working with a faulty model... what about when people actually have a good model, and a disinformation stream replaces that with a bad model, or causes them to just say 'nothing is true'? For your purposes of deep truth, sure, everyone knows almost nothing except maybe in their own tight discipline, so one noise level or another will make no difference. But at the level I'm talking about, common understanding of basic truths (the sky is blue, scattering, not a clue abt quantum effects), we are already seeing how increasing the noise level only by removing the Fairness Doctrine in broadcast media — society is significantly more fragmented. Automating and scaling those drivers of untruths is likely to accelerate the fragmentation.
carlosjobim 3y ago
In the past, being able to discern truth from lies would be the difference between life and death for most individuals. Still today, there are professions and activities, where you will be killed or seriously injured if you don't understand the difference between truth and lies. That is in contrast to a lot of academia, politics, online message boards and other purely inter-personal stuff, where truth is what the majority or the most active group decides it is, because it doesn't matter so much. Separating truth from lies is something every one of us as a human have to do, and that we can't outsource with complete confidence to anybody else. Credentials are one of the last factors that should be weighed when determining truth vs lies.

toss1 3y ago

>>It has never been realistic for people to find the truth, Rather strained definition of truth there. The truth that matters is a basic not-very-wrong version of what is happening. Examples: 1) The earth is roughly spheroidal. 2) NASA is a real space agency and landed people on the moon in 1969-1973. 3) The trails behind airplanes are condensation particles, not sprayed chemicals. 4) Vaccines are safe and effective, and the ratio of issues to prevented disease is on the order of millions-to-one, a vast benefit to society, especially those who medically cannot get vaccines. 5) Hijacked airlines and not pre-set explosives brought town the WTC towers in NYC on 9/11/2001. Etc. Etc. Etc.... We do NOT need a doctorate in Geology, selection into NASA Astronaut Corps, graduate astronomic physics course, PhD-level immunology, or a structural engineer's license to understand any of the above to a useful level. Yet, there is sufficient human-generated noise to convince millions that none of these are true. And you are claiming that adding industrial-scale AI agents to the public 'conversation' at a scale never seen in history will have zero effect? Sure, it might completely fizzle. But the normal and prudent expectation when something comes in and "floods the zone with shit" [0], especially when that shit has the potential to be at a volume that is far beyond anything experienced, is to expect that it will have some effect. What evidence do you have that it will have no effect? In answering, consider another top HN story at the moment "Facebook Is Being Overrun with Stolen, AI-Gen Images That People Think Are Real". [1] [0] Steve Bananon's description of a disinformation tactic [1] https://news.ycombinator.com/item?id=38684309 https://news.ycombinator.com/item?id=38684309

randomdata 3y ago
> to understand any of the above to a useful level. The standard model for electrical current is not true, but it is still useful in the vast, vast, vast majority of cases. True and useful are very different things. The conflation here is erroneous. The average person is never going to understand what is true in any meaningful way, and even those who focus on certain truths are going to be focused on very specific truths, not all truths. There is only so much time in the day. For everything else they have to work with simplified models that may be quite disconnected from reality, but as long as they work for all practical purposes, who cares? > And you are claiming that adding industrial-scale AI agents to the public 'conversation' at a scale never seen in history will have zero effect? Conversation only happens when people realize they are working with a faulty model, so if conversation keeps happening then that means people are realizing that the model they currently have doesn't work. What do you think is going to happen with more noise? If people can't find a usable model in a not-so-noisy environment, do you think they will suddenly be able to find one in a noisier environment? That seems quite unlikely. In reality, the outcome is the same either way. Maybe increased noise will turn more interest to the vetted environments and ignoring the noisy conversations, but if people are already not finding the vetted, the reality is that we are not likely to see change there either.
- toss1 3y ago
  >>The average person is never going to understand what is true in any meaningful way, Agree, but this means we are talking about two different topics. I'm talking at a functional level of truth; a base level of truths commonly understood in a society such that it can maintain cohesion. You're talking about deeply understood truths, like the 2+-hour discussion of quantum effects on the question of "why is the sky blue?" at a PhD's defense presentation. >> Conversation only happens when people realize they are working with a faulty model... what about when people actually have a good model, and a disinformation stream replaces that with a bad model, or causes them to just say 'nothing is true'? For your purposes of deep truth, sure, everyone knows almost nothing except maybe in their own tight discipline, so one noise level or another will make no difference. But at the level I'm talking about, common understanding of basic truths (the sky is blue, scattering, not a clue abt quantum effects), we are already seeing how increasing the noise level only by removing the Fairness Doctrine in broadcast media — society is significantly more fragmented. Automating and scaling those drivers of untruths is likely to accelerate the fragmentation.

carlosjobim 3y ago

In the past, being able to discern truth from lies would be the difference between life and death for most individuals. Still today, there are professions and activities, where you will be killed or seriously injured if you don't understand the difference between truth and lies. That is in contrast to a lot of academia, politics, online message boards and other purely inter-personal stuff, where truth is what the majority or the most active group decides it is, because it doesn't matter so much. Separating truth from lies is something every one of us as a human have to do, and that we can't outsource with complete confidence to anybody else. Credentials are one of the last factors that should be weighed when determining truth vs lies.